- Email: [email protected]
The amount and nature of glutathione transferases in rat liver microsomes determined by immunochemical methods
Volume 160, number 1,2
FEBS 0741
August 1983
The amount and nature of glutathione transferases in rat liver microsomes determined by immunochemical methods Ralf Morgenstern, Department
of Biochemistry,
Claes Guthenberg, Arrhenius Laboratory,
Bengt Mannervik and Joseph W. DePierre University of Stockholm,
S-104 91 Stockholm,
Sweden
Received 12 July 1983 The amount and nature of glutathione transferases in rat liver microsomes were determined using immunological techniques. It was shown that cytosolic glutathione transferase subunits A plus C, and B plus L were present at levels of 2.4 k 0.6 and 1.5 f 0.1 pg/mg microsomal protein, respectively. These levels are IO-times higher than those for non-specific binding of cytosolic components judging from the distribution of lactate dehydrogenase, a cytosolic marker. The possibility that a portion of these glutathione transferases is functionally localized on the endoplasmic reticulum is discussed. A previously described microsomal glutathione transferase which is distinct from the cytosolic enzymes is present in an amount of 3 1 + 6 pg/mg microsomal protein. Glutathione transferase
Microsomal
1. INTRODUCTION The glutathione transferases are a family of enzymes with broad and overlapping substrate specificities (review [ 11). These enzymes are important in the detoxication of numerous potentially toxic, carcinogenic and mutagenic substances [2] and can also function as binding proteins [3]. To date at least 7 different proteins in rat liver cytosol with glutathione transferase activity have been described [4,5]. It was recently found that the major part of the activity is due to 6 dimeric proteins formed by binary combinations of 4 different subunits, named A, B, C and L [5,6]. The glutathione transferases previously known as A, AA, B and C are now referred to as transferases AZ, B2, BL and AC, on the basis of their respective subunit compositions. Accordingly, antibodies raised against transferase ‘B’ (now BL) react with both subunits B and L, and antibodies towards transferase ‘C’ (now AC) react with subunits A and C [6]. Antibodies towards the heterodimers AC and BL recognize the corresponding homodimers with the same titer, with the exception of Lz, which is underestimated by 34 rfr 10% (n = 5, SD)
264
Rat liver
Drug metabolism
by the immunoquantitative methods used. By use of such antibodies it was previously found that about 5% of the protein in the cytosol fraction of rat liver is accounted for by subunits A, B, C, and L of the glutathione transferases [7]. The purification of soluble glutathione transferases from mitochondria has been described [8], but the relationship of these enzymes to the cytosolic transferases is not yet clear. Microsomes demonstrate glutathione transferwith 1-chloro-2,4_dinitrobenzene ase activity (CDNB) at a level which is about 10% of that observed in the cytosol. In [9] it was concluded on the basis of immunological and isoelectric focusing experiments that the microsomal glutathione transferases were identical or at least very similar to their cytosolic counterparts, but might have a stable localization on the endoplasmic reticulum since they could not be removed by various washing procedures (see [lo]). In [ 111, we discovered a new microsomal glutathione transferase activity which can be stimulated several-fold with thiol reagents under conditions where the cytosolic activities are not affected [ 111. We also demonstrated that this activity
00145793/83/$3.00
0
Published by Elsevier Science Publishers B. V. 1983 Federation of European Biochemical Societies
Volume 160, number 1,2
FEBS LETTERS
is localized on the endoplasmic reticulum and is not affected by xenobiotics which induce the cytosolic glutathione transferase [ 101. We purified this new microsomal glutathione transferase in both iV-ethylmaleimide-activated [12] and unactivated forms [13]. The purification factor of 36 indicates that this protein constitutes as much as 3% of the total protein of the endoplasmic reticulum. This microsomal transferase differs from the cytosolic enzymes in terms of immunochemical properties, M,, amino acid composition, substrate specificity, and dependence on detergent for activity [ 12,131. The fact that both a distinct microsomal glutathione transferase and glutathione transferases very similar or identical to the cytosolic enzymes are recovered in rat liver microsomes raises important questions as to which of the cytosolic enzymes are present, how much of the different transferases are there, and whether the presence of the cytosolic transferases can be explained simply by contamination. Here, we use immunochemical methods to quantitate the amounts of the microsomal glutathione transferase and the major cytosolic glutathione transferases present in microsomes . 2. MATERIALS
AND METHODS
Liver microsomes from male Sprague-Dawley rats (180-200 g) were prepared as described previously, except that the microsomes were washed twice with 0.15 Tris-HCl (pH 8) to remove cytosolic contamination [lo]. Microsomal glutathione transferase [12] and cytosolic transferases AC (‘C’) and BL (‘B’) were purified as in [5,7,14] and antisera raised in rabbits by conventional methods. Glutathione transferase activity with l-chloro-2,4dinitrobenzene was measured as in [4]. Protein was measured as in [ 151, sometimes as modified in [ 161, with bovine serum albumin as the standard. Radial immunodiffusion was performed on 8 x 8 cm glass plates on which were cast 8 ml of 1% agar containing 0.125 and 0.2 ml of antisera against glutathione transferase AC and BL, respectively; 1% Triton X-100 and 50 mM Tris-glycine (pH 8.6) (transferase AC) or 50 mM potassium phosphate buffer (pH 7) (transferase BL) were also included. Purified enzymes were used as standards. Wells of 5 mm diameter were punched and
August 1983
20 ~1 of different dilutions of standard proteins or (solubilized with 3 mg Triton microsomes X-lOO/mg protein) was added. The plates were left to stand at room temperature in a humid atmosphere for 48-72 h and after washing in several changes of 1% Triton X-100, 0.9% NaCl and 50 mM potassium phosphate (pH 7) the gels were stained with Coomassie brilliant blue as in [ 171. Rocket immunoelectrophoresis for quantitation of the subunits of glutathione transferase BL was performed on 8 x 8 cm glass plates on which were cast 8 ml 1% agar gel containing 0.25 ml of antisera 23 mM barbiturate buffer (pH 8.6), 1% Triton X-100; 5 ~1 of different dilutions of pure proteins or microsomes were added to the wells as described above. The gel was run overnight at 4°C with 4 V/cm. Gels were washed and stained as above. The amount of microsomal glutathione transferase was determined by immunoblotting [ 181 after sodium dodecylsulphate (SDS)-polyacrylamide gel electrophoresis (15%) as in [ 171. The amount of protein on the blot was quantitated by the use of radioiodinated protein A. Fig.1 shows standard curves for typical experiments with all 3 techniques. All chemicals used were from common commercial sources and of the highest purity available. Radioiodinated protein A was a generous gift from Dr A. Wielburski of this department. 3. RESULTS AND DISCUSSION In this study 3 different immunochemical methods for enzyme quantitation were used. The glutathione transferases of the A-C group [5] have isoelectric points close to immunoglobulins and are measured by radial immunodiffusion. Glutathione transferases of the B-L group have higher isoelectric points and hence can be estimated both by rocket immunoelectrophoresis and radial immunodiffusion. Antisera against the microsomal glutathione transferase also recognize a few microsomal peptides of higher Mr and therefore immunoblotting after separation on SDS-PAGE is a specific way to measure the content of this enzyme. The results summarized in table 1 show that the microsomal glutathione transferase accounts for 3.1% of the total protein in rat liver microsomes 265
Volume 160, number 1,2
FEBS LETTERS
August 1983
Fig la
Fig.lc
Fig lb
Q27
OL
OL , 01
0.54 0.8
0.3
0.2 pg
).Igprote1n
I
proto1n
2
3
pg
protein
L
5
Fig.1. (a) Standard curve for: (a) Radial immunodiffusion - (0) glutathione transferases A2 + AC + C2; (0) glutathione transferases B2 + BL + L2. (b) Rocket immunoelectrophoresis (0) glutathione transferases B2 + BL + I-2. (c) Immunoblotting - (0) microsomal glutathione transferase. The experiments are described in section 2.
Table 1 Amounts of different glutathione Glutathione transferase
transferases
pg/mg Microsomal protein=
in rat liver microsomes
Calculated contribution to microsomal activityb Untreated
Microsomal A2 + AC + CZ Bz + BL + LZ Total
31 + 0.6 (n = 7) 2.4 f 0.6 (n = 4) 1.5 * 0.1 (n = 3) 35
62 (36%) 67 (39%) 41 (24%) 170d
NEM-activatedC 930 (90010) 67 ( 6%) 41 ( 4%) 1038e
a Means + SD from two separate experiments; n is the number of observations that were in the linear portions of standard curves. Transferases A2 + AC + CZ were measured with radial immunodiffusion. Transferases B2 + BL + LZ were also estimated by this method and rocket immunoelectrophoresis. Microsomal glutathione transferase was estimated by immunoblotting b Specific activities (urn01 1-chloro-2,4_dinitrobenzene conjugated. min-’ . mg-‘) used were the mean for the A-C group, 28 and the B-L group, 27 (from [5]) and for unactivated and NEM-activated microsomal transferase 2 and 30, respectively (from 112,131) ’ Activation with N-ethylmaleimide as in [l l] d Observed activity = 120; ’ Observed activity = 800
and is present in great excess over the cytosolic enzymes. The value of 3.1% for the microsomal enzyme agrees well with a purification factor of 36 (2.8% content) [12]. The value for the transferase 266
BL family is somewhat underestimated because the sera have a lower titer with transferase La, Since all glutathione transferases appear to have different substrate specificities [ 1,5] the relative contribu-
Volume 160, number 1,2
FEBSLETTERS
tion of each form to the conjugation of any given substrate with glutathione by microsomes will depend on the nature of the compound. Activation of the microsomal enzyme by thiol reagents can, of course, also change the pattern. In table 1 we have also calculated the relative contribution of the cytosolic and microsomal enzyme forms to the conjugation of l-chloro-2,4dinitrobenzene in microsomes before and after activation. As can be seen, activation shifts the relative contribution of the microsomal enzyme from 36% to 90%. When comparing the values obtained by calculation of the total activity using the contents and specific activities of the individual enzymes, reasonable agreement with the total measurable activity is found. The slightly higher total activity obtained by calculation might reflect the fact that some or all of the enzymes have higher activity in solubilized, isolated form than in situ in the microsomal membrane. This view is supported by the observation that solubilization of microsomes in Triton X-100 gives a slight increase in activity. It is thus clear that microsomes possess the potential for catalyzing the conjugation of glutathione to a variety of compounds. The location of such activity in the endoplasmic reticulum might be advantageous for two reasons: (i) Most substrates for glutathione transferases are lipophilic and are expected to dissolve extensively in membranes; (ii) Electrophilic metabolites from the cytochrome P-450 system are generated in this membrane system and can be substrates for these enzymes (see [19]). Work now performed in our laboratory demonstrates that the microsomal glutathione transferases have a broad substrate specificity [ 131, in analogy to the cytosolic enzymes [l]. The question as to whether the ‘cytosolic’ glutathione transferases are actually functionally localized on the endoplasmic reticulum cannot be answered conclusively at present. Evidence from different washing procedures [9, lo] suggests that the presence of these enzymes in microsomes does not simply reflect cytosolic contamination. In addition, comparison with the level of the cytosolic marker lactate dehydrogenase in microsomes reveals that the ‘cytosolic’ glutathione transferases are present at a lo-times higher level than would be
August 1983
expected on the basis of contamination [9]. Furthermore, induction of the cytosolic activity up to 3-4-fold does not increase the microsomal activity [lo]. This fact speaks against contamination and for the presence of defined amounts of the cytosolic enzymes in relationship to microsomal protein. The situation in mouse liver microsomes is the same. Authors in [20] have identified the soluble enzyme FZ and FJ as associated to the microsomal membrane by immunological and other methods. However, their work was not concerned with the microsomal glutathione transferase activatable by thiol reagents which is also present in mouse microsomes [21]. Evidence from immunoblotting experiments (not shown) reveals that a form in the mouse, which has the same M,, crossreacts with antibodies against the rat enzyme. It is reasonable to assume from the similarity in activity after activation that the microsomal form is present in large excess, also in the mouse. If the ‘cytosolic’ glutathione transferases are stable components of the endoplasmic reticulum in vivo, they may be found to differ slightly (e.g., in molecular mass or amino acid composition) from the soluble enzymes. Another possibility is that cytosolic glutathione transferases may bind to the endoplasmic reticulum under certain conditions and be released under other conditions, as is the case with hexokinase and mitochondria in rat brain [22]. We have observed that removal of ribosomes does not decrease microsomal glutathione transferase activity indicating that the microsomal ‘cytosolic’ glutathione transferases are probably not newly-synthesized proteins which have not yet been released into the cytoplasm. We have here examined the major forms of cytosolic glutathione transferase, which, together with the microsomal enzyme, seem to account for all of the conjugating activity towards l-chloro2,4-dinitrobenzene found in microsomes (table 1). However, it is possible that small amounts of the minor soluble forms may also be present in microsomes. The high level of the microsomal glutathione transferase suggests that, like the cytosolic transferase, it might also function as a binding protein. Another possible role for this enzyme is suggested by the finding [23-251 that there is a heat-labile factor in microsomes which can protect against lipid peroxidation. The microsomal glutathione 267
Volume 160, number 1,2
FBBSLETTERS
transferase might be involved here, since microsomal glutathione peroxidase activity towards cumene hydroperoxide can be activated 2-fold by N-ethylmaleimide [23].We have found that the purified microsomal glutathione transferase has glutathione peroxidase activity towards cumene hydroperoxide which can be activated IO-fold by N-ethylmaleimide [ 131. ACKNOWLEDGEMENTS These &dies were supported by National Institute of Health grant 1 RO 1 CA 26261-02 awarded by the National Cancer Institute (Department of Health, Education and Welfare, Bethesda MD), by the Swedish National Science Research Council and the Swedish Cancer Society. The generous support of R.M. from the P.E. Lindahls Fond is gratefully acknowledged. We wish to thank Kerstin Larsen, Majken Brandt and S. Svensson for preparing the antisera and Helgi Jensson for providing purified soluble glutathione transferases. REFERENCES 111 Jakoby,
W.B. and Habig, W.H. (1980) in: Enzymatic Basis of Detoxication (Jakoby, W.B. ed) ~01.2, pp.63-94, Academic Press, New York. PI Chasseaud, L.F. (1979) Adv. Cancer Res. 29, 175-274. 131 Smith, G.J. and Litwack, G. (1980) Rev. Biochem. Toxicol. 2, l-47. 141 Habig, W.H., Pabst, M.J. and Jakoby, W.B. (1974) j. Biol. Chem. 249, 7130-7139. 151 Mannervik, B. and Jensson, H. (1982) J. Biol. Chem. 257, 9909-9912. B., Guthenberg, C., Jensson, H., [61 Mannervik, Warholm, M. and Alin, P. (1983) in: Functions of Glutathione Biochemical, Physiological, Clinical and Toxicological Aspects (Larsson, A. et al. eds) pp.75-88, Raven Press, New York.
268
August 1983
[71 Guthenberg, C., Morgenstern, R., DePierre, J.W. and Mannervik, B. (1980) Biochim. Biophys. Acta 631, l-10. [81 Kraus, P. (1980) Hoppe-Seyler’s Z. Physiol. Chem. 361, 9-15. [91 Friedberg, T., Bentley, P., Stasiecki, P., Glatt, H.R., Raphael, D. and Oesch, F. (1979) J. Biol. Chem. 254, 12028-12033. UOI Morgenstern, R., Meijer, J., DePierre, J.W. and Ernster, L. (1980) Eur. J. Biochem. 104, 167-174. illI Morgenstern, R., DePierre, J.W. and Ernster, L. (1979) Biochem. Biophys. Res. Commun. 87, 657-663. 1121 Morgenstern, R., Guthenberg, C. and DePierre, J.W. (1982) Eur. J. Biochem. 128, 243-248. 1131 Morgenstern, R. and DePierre, J.W. (1983) Eur. J. Biochem., in press. u41 Askelbf, P., Guthenberg, C., Jakobson, I. and Mannervik, B. (1975) Biochem. J. 147, 513-522. v51 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. G.L. (1977) Anal. Biochem. 83, 1161 Peterson, 346-356. 1171 Laemmli, U.K. (1970) Nature 227, 680-685. WI Towbin, H., Staehelin, T. and Gordon, L. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. I191 Morgenstern, R., Guthenberg, C., Mannervik, B., DePierre, J.W. and Ernster, L. (1982) Cancer Res. 42, 4215-4221. [201 Lee, G.C.-Y. and McKinney, J.D. (1982) Mol. Cell Biochem. 48, 91-96. [211 Morgenstern, R. and Dock, L. (1982) Acta Chem. Stand. B36, 255-256. 1221 Rose, I.A. and Warms, J.W.B. (1967) J. Biol. Chem. 242, 1635-1645. [231 Reddy, C.C., Tu, C.-P.D., Burgess, J.R., Ho, C.Y., Scholz, R.W. and Massaro, E.J. (1981) Biothem. Bioohvs. Res. Commun. 101, 970-978. Reddy, CC, Scholz, R.W., Thomas, C.E. and WI Massaro, E.J. (1982) Life Sci. 31, 571-576. 31, i251 Burk, R.F. (1982) Biochem. Pharmacol. 601-602:
FEBS 0741
August 1983
The amount and nature of glutathione transferases in rat liver microsomes determined by immunochemical methods Ralf Morgenstern, Department
of Biochemistry,
Claes Guthenberg, Arrhenius Laboratory,
Bengt Mannervik and Joseph W. DePierre University of Stockholm,
S-104 91 Stockholm,
Sweden
Received 12 July 1983 The amount and nature of glutathione transferases in rat liver microsomes were determined using immunological techniques. It was shown that cytosolic glutathione transferase subunits A plus C, and B plus L were present at levels of 2.4 k 0.6 and 1.5 f 0.1 pg/mg microsomal protein, respectively. These levels are IO-times higher than those for non-specific binding of cytosolic components judging from the distribution of lactate dehydrogenase, a cytosolic marker. The possibility that a portion of these glutathione transferases is functionally localized on the endoplasmic reticulum is discussed. A previously described microsomal glutathione transferase which is distinct from the cytosolic enzymes is present in an amount of 3 1 + 6 pg/mg microsomal protein. Glutathione transferase
Microsomal
1. INTRODUCTION The glutathione transferases are a family of enzymes with broad and overlapping substrate specificities (review [ 11). These enzymes are important in the detoxication of numerous potentially toxic, carcinogenic and mutagenic substances [2] and can also function as binding proteins [3]. To date at least 7 different proteins in rat liver cytosol with glutathione transferase activity have been described [4,5]. It was recently found that the major part of the activity is due to 6 dimeric proteins formed by binary combinations of 4 different subunits, named A, B, C and L [5,6]. The glutathione transferases previously known as A, AA, B and C are now referred to as transferases AZ, B2, BL and AC, on the basis of their respective subunit compositions. Accordingly, antibodies raised against transferase ‘B’ (now BL) react with both subunits B and L, and antibodies towards transferase ‘C’ (now AC) react with subunits A and C [6]. Antibodies towards the heterodimers AC and BL recognize the corresponding homodimers with the same titer, with the exception of Lz, which is underestimated by 34 rfr 10% (n = 5, SD)
264
Rat liver
Drug metabolism
by the immunoquantitative methods used. By use of such antibodies it was previously found that about 5% of the protein in the cytosol fraction of rat liver is accounted for by subunits A, B, C, and L of the glutathione transferases [7]. The purification of soluble glutathione transferases from mitochondria has been described [8], but the relationship of these enzymes to the cytosolic transferases is not yet clear. Microsomes demonstrate glutathione transferwith 1-chloro-2,4_dinitrobenzene ase activity (CDNB) at a level which is about 10% of that observed in the cytosol. In [9] it was concluded on the basis of immunological and isoelectric focusing experiments that the microsomal glutathione transferases were identical or at least very similar to their cytosolic counterparts, but might have a stable localization on the endoplasmic reticulum since they could not be removed by various washing procedures (see [lo]). In [ 111, we discovered a new microsomal glutathione transferase activity which can be stimulated several-fold with thiol reagents under conditions where the cytosolic activities are not affected [ 111. We also demonstrated that this activity
00145793/83/$3.00
0
Published by Elsevier Science Publishers B. V. 1983 Federation of European Biochemical Societies
Volume 160, number 1,2
FEBS LETTERS
is localized on the endoplasmic reticulum and is not affected by xenobiotics which induce the cytosolic glutathione transferase [ 101. We purified this new microsomal glutathione transferase in both iV-ethylmaleimide-activated [12] and unactivated forms [13]. The purification factor of 36 indicates that this protein constitutes as much as 3% of the total protein of the endoplasmic reticulum. This microsomal transferase differs from the cytosolic enzymes in terms of immunochemical properties, M,, amino acid composition, substrate specificity, and dependence on detergent for activity [ 12,131. The fact that both a distinct microsomal glutathione transferase and glutathione transferases very similar or identical to the cytosolic enzymes are recovered in rat liver microsomes raises important questions as to which of the cytosolic enzymes are present, how much of the different transferases are there, and whether the presence of the cytosolic transferases can be explained simply by contamination. Here, we use immunochemical methods to quantitate the amounts of the microsomal glutathione transferase and the major cytosolic glutathione transferases present in microsomes . 2. MATERIALS
AND METHODS
Liver microsomes from male Sprague-Dawley rats (180-200 g) were prepared as described previously, except that the microsomes were washed twice with 0.15 Tris-HCl (pH 8) to remove cytosolic contamination [lo]. Microsomal glutathione transferase [12] and cytosolic transferases AC (‘C’) and BL (‘B’) were purified as in [5,7,14] and antisera raised in rabbits by conventional methods. Glutathione transferase activity with l-chloro-2,4dinitrobenzene was measured as in [4]. Protein was measured as in [ 151, sometimes as modified in [ 161, with bovine serum albumin as the standard. Radial immunodiffusion was performed on 8 x 8 cm glass plates on which were cast 8 ml of 1% agar containing 0.125 and 0.2 ml of antisera against glutathione transferase AC and BL, respectively; 1% Triton X-100 and 50 mM Tris-glycine (pH 8.6) (transferase AC) or 50 mM potassium phosphate buffer (pH 7) (transferase BL) were also included. Purified enzymes were used as standards. Wells of 5 mm diameter were punched and
August 1983
20 ~1 of different dilutions of standard proteins or (solubilized with 3 mg Triton microsomes X-lOO/mg protein) was added. The plates were left to stand at room temperature in a humid atmosphere for 48-72 h and after washing in several changes of 1% Triton X-100, 0.9% NaCl and 50 mM potassium phosphate (pH 7) the gels were stained with Coomassie brilliant blue as in [ 171. Rocket immunoelectrophoresis for quantitation of the subunits of glutathione transferase BL was performed on 8 x 8 cm glass plates on which were cast 8 ml 1% agar gel containing 0.25 ml of antisera 23 mM barbiturate buffer (pH 8.6), 1% Triton X-100; 5 ~1 of different dilutions of pure proteins or microsomes were added to the wells as described above. The gel was run overnight at 4°C with 4 V/cm. Gels were washed and stained as above. The amount of microsomal glutathione transferase was determined by immunoblotting [ 181 after sodium dodecylsulphate (SDS)-polyacrylamide gel electrophoresis (15%) as in [ 171. The amount of protein on the blot was quantitated by the use of radioiodinated protein A. Fig.1 shows standard curves for typical experiments with all 3 techniques. All chemicals used were from common commercial sources and of the highest purity available. Radioiodinated protein A was a generous gift from Dr A. Wielburski of this department. 3. RESULTS AND DISCUSSION In this study 3 different immunochemical methods for enzyme quantitation were used. The glutathione transferases of the A-C group [5] have isoelectric points close to immunoglobulins and are measured by radial immunodiffusion. Glutathione transferases of the B-L group have higher isoelectric points and hence can be estimated both by rocket immunoelectrophoresis and radial immunodiffusion. Antisera against the microsomal glutathione transferase also recognize a few microsomal peptides of higher Mr and therefore immunoblotting after separation on SDS-PAGE is a specific way to measure the content of this enzyme. The results summarized in table 1 show that the microsomal glutathione transferase accounts for 3.1% of the total protein in rat liver microsomes 265
Volume 160, number 1,2
FEBS LETTERS
August 1983
Fig la
Fig.lc
Fig lb
Q27
OL
OL , 01
0.54 0.8
0.3
0.2 pg
).Igprote1n
I
proto1n
2
3
pg
protein
L
5
Fig.1. (a) Standard curve for: (a) Radial immunodiffusion - (0) glutathione transferases A2 + AC + C2; (0) glutathione transferases B2 + BL + L2. (b) Rocket immunoelectrophoresis (0) glutathione transferases B2 + BL + I-2. (c) Immunoblotting - (0) microsomal glutathione transferase. The experiments are described in section 2.
Table 1 Amounts of different glutathione Glutathione transferase
transferases
pg/mg Microsomal protein=
in rat liver microsomes
Calculated contribution to microsomal activityb Untreated
Microsomal A2 + AC + CZ Bz + BL + LZ Total
31 + 0.6 (n = 7) 2.4 f 0.6 (n = 4) 1.5 * 0.1 (n = 3) 35
62 (36%) 67 (39%) 41 (24%) 170d
NEM-activatedC 930 (90010) 67 ( 6%) 41 ( 4%) 1038e
a Means + SD from two separate experiments; n is the number of observations that were in the linear portions of standard curves. Transferases A2 + AC + CZ were measured with radial immunodiffusion. Transferases B2 + BL + LZ were also estimated by this method and rocket immunoelectrophoresis. Microsomal glutathione transferase was estimated by immunoblotting b Specific activities (urn01 1-chloro-2,4_dinitrobenzene conjugated. min-’ . mg-‘) used were the mean for the A-C group, 28 and the B-L group, 27 (from [5]) and for unactivated and NEM-activated microsomal transferase 2 and 30, respectively (from 112,131) ’ Activation with N-ethylmaleimide as in [l l] d Observed activity = 120; ’ Observed activity = 800
and is present in great excess over the cytosolic enzymes. The value of 3.1% for the microsomal enzyme agrees well with a purification factor of 36 (2.8% content) [12]. The value for the transferase 266
BL family is somewhat underestimated because the sera have a lower titer with transferase La, Since all glutathione transferases appear to have different substrate specificities [ 1,5] the relative contribu-
Volume 160, number 1,2
FEBSLETTERS
tion of each form to the conjugation of any given substrate with glutathione by microsomes will depend on the nature of the compound. Activation of the microsomal enzyme by thiol reagents can, of course, also change the pattern. In table 1 we have also calculated the relative contribution of the cytosolic and microsomal enzyme forms to the conjugation of l-chloro-2,4dinitrobenzene in microsomes before and after activation. As can be seen, activation shifts the relative contribution of the microsomal enzyme from 36% to 90%. When comparing the values obtained by calculation of the total activity using the contents and specific activities of the individual enzymes, reasonable agreement with the total measurable activity is found. The slightly higher total activity obtained by calculation might reflect the fact that some or all of the enzymes have higher activity in solubilized, isolated form than in situ in the microsomal membrane. This view is supported by the observation that solubilization of microsomes in Triton X-100 gives a slight increase in activity. It is thus clear that microsomes possess the potential for catalyzing the conjugation of glutathione to a variety of compounds. The location of such activity in the endoplasmic reticulum might be advantageous for two reasons: (i) Most substrates for glutathione transferases are lipophilic and are expected to dissolve extensively in membranes; (ii) Electrophilic metabolites from the cytochrome P-450 system are generated in this membrane system and can be substrates for these enzymes (see [19]). Work now performed in our laboratory demonstrates that the microsomal glutathione transferases have a broad substrate specificity [ 131, in analogy to the cytosolic enzymes [l]. The question as to whether the ‘cytosolic’ glutathione transferases are actually functionally localized on the endoplasmic reticulum cannot be answered conclusively at present. Evidence from different washing procedures [9, lo] suggests that the presence of these enzymes in microsomes does not simply reflect cytosolic contamination. In addition, comparison with the level of the cytosolic marker lactate dehydrogenase in microsomes reveals that the ‘cytosolic’ glutathione transferases are present at a lo-times higher level than would be
August 1983
expected on the basis of contamination [9]. Furthermore, induction of the cytosolic activity up to 3-4-fold does not increase the microsomal activity [lo]. This fact speaks against contamination and for the presence of defined amounts of the cytosolic enzymes in relationship to microsomal protein. The situation in mouse liver microsomes is the same. Authors in [20] have identified the soluble enzyme FZ and FJ as associated to the microsomal membrane by immunological and other methods. However, their work was not concerned with the microsomal glutathione transferase activatable by thiol reagents which is also present in mouse microsomes [21]. Evidence from immunoblotting experiments (not shown) reveals that a form in the mouse, which has the same M,, crossreacts with antibodies against the rat enzyme. It is reasonable to assume from the similarity in activity after activation that the microsomal form is present in large excess, also in the mouse. If the ‘cytosolic’ glutathione transferases are stable components of the endoplasmic reticulum in vivo, they may be found to differ slightly (e.g., in molecular mass or amino acid composition) from the soluble enzymes. Another possibility is that cytosolic glutathione transferases may bind to the endoplasmic reticulum under certain conditions and be released under other conditions, as is the case with hexokinase and mitochondria in rat brain [22]. We have observed that removal of ribosomes does not decrease microsomal glutathione transferase activity indicating that the microsomal ‘cytosolic’ glutathione transferases are probably not newly-synthesized proteins which have not yet been released into the cytoplasm. We have here examined the major forms of cytosolic glutathione transferase, which, together with the microsomal enzyme, seem to account for all of the conjugating activity towards l-chloro2,4-dinitrobenzene found in microsomes (table 1). However, it is possible that small amounts of the minor soluble forms may also be present in microsomes. The high level of the microsomal glutathione transferase suggests that, like the cytosolic transferase, it might also function as a binding protein. Another possible role for this enzyme is suggested by the finding [23-251 that there is a heat-labile factor in microsomes which can protect against lipid peroxidation. The microsomal glutathione 267
Volume 160, number 1,2
FBBSLETTERS
transferase might be involved here, since microsomal glutathione peroxidase activity towards cumene hydroperoxide can be activated 2-fold by N-ethylmaleimide [23].We have found that the purified microsomal glutathione transferase has glutathione peroxidase activity towards cumene hydroperoxide which can be activated IO-fold by N-ethylmaleimide [ 131. ACKNOWLEDGEMENTS These &dies were supported by National Institute of Health grant 1 RO 1 CA 26261-02 awarded by the National Cancer Institute (Department of Health, Education and Welfare, Bethesda MD), by the Swedish National Science Research Council and the Swedish Cancer Society. The generous support of R.M. from the P.E. Lindahls Fond is gratefully acknowledged. We wish to thank Kerstin Larsen, Majken Brandt and S. Svensson for preparing the antisera and Helgi Jensson for providing purified soluble glutathione transferases. REFERENCES 111 Jakoby,
W.B. and Habig, W.H. (1980) in: Enzymatic Basis of Detoxication (Jakoby, W.B. ed) ~01.2, pp.63-94, Academic Press, New York. PI Chasseaud, L.F. (1979) Adv. Cancer Res. 29, 175-274. 131 Smith, G.J. and Litwack, G. (1980) Rev. Biochem. Toxicol. 2, l-47. 141 Habig, W.H., Pabst, M.J. and Jakoby, W.B. (1974) j. Biol. Chem. 249, 7130-7139. 151 Mannervik, B. and Jensson, H. (1982) J. Biol. Chem. 257, 9909-9912. B., Guthenberg, C., Jensson, H., [61 Mannervik, Warholm, M. and Alin, P. (1983) in: Functions of Glutathione Biochemical, Physiological, Clinical and Toxicological Aspects (Larsson, A. et al. eds) pp.75-88, Raven Press, New York.
268
August 1983
[71 Guthenberg, C., Morgenstern, R., DePierre, J.W. and Mannervik, B. (1980) Biochim. Biophys. Acta 631, l-10. [81 Kraus, P. (1980) Hoppe-Seyler’s Z. Physiol. Chem. 361, 9-15. [91 Friedberg, T., Bentley, P., Stasiecki, P., Glatt, H.R., Raphael, D. and Oesch, F. (1979) J. Biol. Chem. 254, 12028-12033. UOI Morgenstern, R., Meijer, J., DePierre, J.W. and Ernster, L. (1980) Eur. J. Biochem. 104, 167-174. illI Morgenstern, R., DePierre, J.W. and Ernster, L. (1979) Biochem. Biophys. Res. Commun. 87, 657-663. 1121 Morgenstern, R., Guthenberg, C. and DePierre, J.W. (1982) Eur. J. Biochem. 128, 243-248. 1131 Morgenstern, R. and DePierre, J.W. (1983) Eur. J. Biochem., in press. u41 Askelbf, P., Guthenberg, C., Jakobson, I. and Mannervik, B. (1975) Biochem. J. 147, 513-522. v51 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. G.L. (1977) Anal. Biochem. 83, 1161 Peterson, 346-356. 1171 Laemmli, U.K. (1970) Nature 227, 680-685. WI Towbin, H., Staehelin, T. and Gordon, L. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. I191 Morgenstern, R., Guthenberg, C., Mannervik, B., DePierre, J.W. and Ernster, L. (1982) Cancer Res. 42, 4215-4221. [201 Lee, G.C.-Y. and McKinney, J.D. (1982) Mol. Cell Biochem. 48, 91-96. [211 Morgenstern, R. and Dock, L. (1982) Acta Chem. Stand. B36, 255-256. 1221 Rose, I.A. and Warms, J.W.B. (1967) J. Biol. Chem. 242, 1635-1645. [231 Reddy, C.C., Tu, C.-P.D., Burgess, J.R., Ho, C.Y., Scholz, R.W. and Massaro, E.J. (1981) Biothem. Bioohvs. Res. Commun. 101, 970-978. Reddy, CC, Scholz, R.W., Thomas, C.E. and WI Massaro, E.J. (1982) Life Sci. 31, 571-576. 31, i251 Burk, R.F. (1982) Biochem. Pharmacol. 601-602: